SLA Impact Calculator — SLA/SLO-based Impact Assessment Guide
A framework for quantitatively measuring and reporting the business impact of incidents.
SLA/SLO/SLI Framework
SLI (Service Level Indicator): Measurement metric
e.g.: Request success rate = Successful requests / Total requests
SLO (Service Level Objective): Internal target
e.g.: Request success rate >= 99.95% (30-day basis)
SLA (Service Level Agreement): External contract
e.g.: Availability < 99.9% -> 10% credit refund on monthly fee
Allowed Downtime by Availability Level
| Availability |
Annual Downtime |
Monthly |
Weekly |
Daily |
| 99% |
3d 15h |
7h 18m |
1h 41m |
14m 24s |
| 99.5% |
1d 19h |
3h 39m |
50m 24s |
7m 12s |
| 99.9% |
8h 46m |
43m 50s |
10m 5s |
1m 26s |
| 99.95% |
4h 23m |
21m 55s |
5m 2s |
43s |
| 99.99% |
52m 36s |
4m 23s |
1m |
8.6s |
| 99.999% |
5m 16s |
26s |
6s |
0.86s |
Error Budget Consumption Calculation
Monthly error budget = (1 - SLO) x 30 days x 24 hours x 60 minutes
Example: SLO = 99.95%
Error budget = 0.0005 x 30 x 24 x 60 = 21.6 minutes/month
When a 30-minute incident occurs:
+-- Budget exceeded: 30 min - 21.6 min = 8.4 min over
+-- Consumption rate: 30 / 21.6 = 138.9% (exceeded!)
+-- Remaining budget: -8.4 min (no additional incidents allowed for the rest of the period)
Business Impact Matrix
Revenue Loss Estimation
Per-minute revenue loss = Daily average revenue / (24 x 60)
Example: Daily average revenue = $100,000
Per-minute revenue loss = 100,000 / 1,440 = $69.44/min
30-minute incident:
+-- Direct loss: $69.44 x 30 = $2,083
+-- Peak hour adjustment: x 2.5 (if during peak hours)
| -> $5,208
+-- Churn estimate (5% of affected customers churn):
| Affected customers x 5% x Customer Lifetime Value (CLV)
+-- Total estimated loss: Direct + Churn + Recovery costs
Impact Severity Levels
| Level |
Criteria |
Example |
Escalation |
| SEV-1 (Critical) |
Core service complete outage |
Payments down, entire site down |
Immediate -> C-level |
| SEV-2 (Major) |
Core feature partial outage |
Specific payment method unavailable |
Within 30 min -> VP |
| SEV-3 (Minor) |
Non-core feature outage |
Recommendation system malfunction |
Within 1 hour -> Manager |
| SEV-4 (Low) |
User experience degradation |
Slow page loading |
Next business day |
User Impact Scope Calculation
Estimating affected users:
1. Normal traffic for the incident timeframe: DAU x (incident_hours/24)
2. Based on error logs: Unique users receiving 5xx responses
3. Feature usage rate: DAU percentage for the affected feature
Example:
DAU = 100,000
Incident time = 14:00-14:30 (peak)
Time window ratio = 8% (peak hour adjustment)
Affected users = 100,000 x 0.08 x (30/60) = 4,000
Financial Impact of SLA Violations
SLA penalty calculation:
Basic structure:
+-- Tier 1: Availability < SLA by 0.1%p -> 10% credit on monthly fee
+-- Tier 2: Availability < SLA by 1.0%p -> 25% credit on monthly fee
+-- Tier 3: Availability < SLA by 5.0%p -> 50% credit on monthly fee
Example: Monthly contract $10,000, SLA 99.9%, actual 99.5%
Shortfall: 0.4%p -> Tier 1
Credit: $10,000 x 10% = $1,000
Composite SLA Calculation
Serial system: SLA_total = SLA_A x SLA_B x SLA_C
e.g.: 99.9% x 99.9% x 99.9% = 99.7% (0.3% = 2h 11m/month)
Parallel system: SLA_total = 1 - (1-SLA_A) x (1-SLA_B)
e.g.: 1 - (0.001 x 0.001) = 99.9999%
Real system:
API Gateway(99.99%) -> [App Server(99.95%) || App Server(99.95%)] -> DB(99.99%)
= 99.99% x (1-(0.0005)^2) x 99.99%
= 99.99% x 99.99975% x 99.99%
~ 99.97%
Report Template
# Incident Impact Assessment Report
## Summary
| Item | Value |
|------|-------|
| Incident Time | YYYY-MM-DD HH:MM ~ HH:MM (N min) |
| Severity Level | SEV-N |
| Affected Users | Approx. N |
| SLA Violation | Yes/No |
| Error Budget Consumption | N% |
| Estimated Revenue Loss | $N |
## Detailed Assessment
### SLI/SLO Impact
### User Impact
### Financial Impact
### Reputation Impact
## Recovery Cost
- Engineering hours: N person-hours
- Infrastructure cost: $N
- External vendor cost: $N
revfactory